This document compares outcomes of anterior cervical discectomy fusion (ACDF) and anterior cervical corpectomy fusion (ACCF) for treating multi-level cervical spondylosis. It retrospectively analyzed 80 patients who underwent either multi-level ACDF (42 patients) or single/multi-level ACCF (38 patients) using titanium mesh cages filled with autograft and anterior cervical plates. The ACCF group had significantly higher blood loss and lower operative time than the ACDF group. Both groups showed equivalent improvement in symptoms, with fusion rates over 90% and maintained cervical lordosis. However, the ACCF group had a higher rate of early hardware failures and pseudarthroses.

1. ORIGINAL ARTICLE
Comparison between anterior cervical discectomy fusion
and cervical corpectomy fusion using titanium cages
for reconstruction: analysis of outcome and long-term follow-up
Juan S. Uribe Æ Jaypal Reddy Sangala Æ
Edward A.M. Duckworth Æ Fernando L. Vale
Received: 2 July 2008 / Accepted: 19 January 2009 / Published online: 12 February 2009
Ó Springer-Verlag 2009
Abstract Retrospective comparative study of 80 con-
secutive patients treated with either anterior cervical
discectomy fusion (ACDF) or anterior cervical corpectomy
fusion (ACCF) for multi-level cervical spondylosis.
To compare clinical outcome, fusion rates, and compli-
cations of anterior cervical reconstruction of multi-level
ACDF and single-/multi-level ACCF performed using
titanium mesh cages (TMCs) filled with autograft and
anterior cervical plates (ACPs). Reconstruction of the
cervical spine after discectomy or corpectomy with tita-
nium cages filled with autograft has become an acceptable
alternative to both allograft and autograft; however, there
is no data comparing the outcome of multi-level ACDF
and single-/multi-level ACCF using this reconstruction.
We evaluated 80 consecutive patients who underwent
surgery for the treatment of multi-level cervical spondy-
losis at our institution from 1998 to 2001. In this series,
42 patients underwent multi-level ACDF (Group 1) and
38 patients underwent ACCF (Group 2). Interbody TMCs
and local autograft bone with ACPs were used in both
procedures. Medical records were reviewed to assess
outcome. Clinical outcome was measured by Odom’s
criteria. Operative time and blood loss were noted.
Radiographs were obtained at 6 and 12 weeks, 6 months,
1 year, and 2 years (if necessary). Early hardware failures
and pseudarthroses were noted. Cervical sagittal curvature
was measured by Ishihara’s index at 1 year. Group 1 had
a mean age 46.2 years (range 35–60 years). Group 2 had
a mean age 50.1 years (range 35–70 years).The operative
time was significantly lower (P 0.001) and blood loss
significantly higher (P 0.001) in Group 2 than in Group
1. At a minimum of 1 year follow up, patients in both
groups had equivalent improvement in their clinical
symptoms. The fusion rates for Group 1 were 97.6 and
92.1% for Group 2. The rates of early hardware failure
were higher in Group 2 (2.6%) than in Group 1 (0%). The
fusion rates for Group 1 were not significantly higher than
Group 2 (P [ 0.28). There was one patient in Group 1
and 2 patients in Group 2 with pseudarthroses. Compli-
cation rates in Group 2 were not significantly higher
(P [ 0.341). Cervical lordosis was well-maintained (80%)
in both groups. Both multi-level ACDF and ACCF with
anterior cervical reconstruction using TMC filled with
autograft and ACP for treatment of multi-level cervical
spondylosis have high fusion rates and good clinical
outcome. However, there is a higher rate of early hard-
ware failure and pseudarthroses after ACCF than ACDF.
Hence, in the absence of specific pathology requiring
removal of vertebral body, multi-level ACDF using
interbody cages and autologous bone graft could result in
lower morbidity.
Keywords Cervical fusion Á Cervical spondylosis Á
Anterior cervical discectomy and fusion Á Anterior
cervical corpectomy
Introduction
Multi-level cervical spondylosis is a challenging and a
common clinical problem [48]. The ideal way of
decompressing the neural elements and reconstructing the
anterior cervical spine in treatment of multi-level
J. S. Uribe Á J. R. Sangala Á E. A.M. Duckworth Á F. L. Vale (&)
Department of Neurosurgery, University of South Florida,
2 Tampa General Circle, USF Health South Center,
Tampa, FL 33606, USA
e-mail: fvale@health.usf.edu
123
Eur Spine J (2009) 18:654–662
DOI 10.1007/s00586-009-0897-9

2. cervical spondylosis is not clear. Although fusion rates
are high with the use of autograft, it is associated with
significant graft site morbidity [43]. Similarly, the use of
allograft, which is devoid of any donor site problems, is
associated with high rates of pseudarthroses [3]. To
overcome these problems associated with both allograft
and autograft, titanium mesh cages (TMCs) have become
popular for anterior cervical reconstruction. The potential
advantages of using interbody cages for reconstruction
after anterior cervical corpectomy fusion (ACCF) include
immediate anterior column stability, avoidance of mor-
bidity associated with autologous bone graft (iliac crest)
harvesting, and good biocompatibility [29]. Similarly, the
use of anterior cervical plating (ACP) after multi-level
anterior cervical discectomy fusion (ACDF) and ACCF
has been regarded as standard practice [58].
It is also not clear whether multi-level cervical spon-
dylosis is best treated with multi-level ACDF or single-/
multi-level ACCF. When compared with multi-level
ACDF (in which there are two surfaces per level), one
hypothesis posits that cervical corpectomy should result in
higher fusion rates because there are only two fusion sur-
faces [14, 30]. However, ACCF has been associated with
early hardware failure [41, 43].
Because of the lack of data comparing the outcome of
multi-level ACDF with ACCF using reconstruction with
TMCs and ACPs, we decided to conduct this retrospective
study. To the best of our knowledge, this is the only study
reported in the English language comparing the outcome of
multi-level ACDF and ACCF with reconstruction using
cages. Similar studies were conducted earlier using allo-
graft and/or autograft (Table 1).
Table 1 Comparative studies of anterior cervical decompression and reconstruction for multi-level cervical spondylosis
Author Decompression Reconstruction Clinical outcome Fusion rates Hardware failures
Yonenobu et al.
[61]
Multi-ACDF vs corpectomy Autograft Better with
corpectomy
Better with
corpectomy
None reported
Brown et al. [6] ACDF Autograft vs allograft Equivalent Equivalent None reported
Fernyhough
et al. [16]
Multi-level ACDF and ACCF Autograft vs allograft Equivalent Better with
autograft
None reported
Bishop et al. [2] Single and multi-level ACDF Autograft vs allograft Autograft
superior
Autograft
superior
None reported
Samartzis et al.
[40]
Multi-level ACDF Autograft vs allograft with
ACP
Equivalent Equivalent Equivalent in both
groups
Rish et al. [38] Multi-level ACDF Autograft vs allograft Equivalent Equivalent None reported
An et al. [1] Multi-level ACDF Autograft vs allograft and
DBX
Autograft
superior
Autograft
superior
None reported
Cauthen et al.
[7]
Multi-level Clowards vs
instrumented Clowards
Autograft vs allograft Autograft
superior
Autograft
superior
Higher in instrumented
Clowards
Moreland et al.
[32]
Single and multi-level ACDF TMC vs allograft Equivalent at
6 months
Equivalent at
6 months
Equivalent at 6 months
Hwang et al.
[23]
Single and multi-level ACDF TMC vs plated TMC Equivalent Equivalent at
12 months
None reported
Swank et al.
[48]
ACDF vs corpectomy Allograft and autograft Corpectomy
group better
Corpectomy
group better
Equivalent
Nirala et al. [33] ACDF vs corpectomy Autograft Corpectomy
group better
Corpectomy
group better
None reported
Cauthen et al.
[8]
ACDF Cage vs dowel vs dowel-
plate
Equivalent Cage group
better
Higher in dowel-plate
Wang et al. [59] Multi-level ACDF vs corpectomy Allograft Equivalent Equivalent Equivalent
Thome et al.
[51]
ACDF Titanium rectangular cages
vs autograft
Equivalent Equivalent Equivalent
Hacker et al.
[22]
ACDF Cages vs uninstrumented
allograft
Equivalent Equivalent None reported
Brazenor [5] Corpectomy Allograft vs titanium rods
and buttress
Equivalent Equivalent None reported
Lind et al. [27] ACDF Titanium cages vs autograft Superior in cage
group
Equivalent None
ACCF anterior cervical corpectomy fusion, ACDF anterior cervical discectomy fusion, TMC titanium mesh cage
Eur Spine J (2009) 18:654–662 655
123

3. Patients and methods
We conducted a retrospective analysis of 80 consecutive
patients treated for multi-level cervical spondylosis
between the years 1998–2001. Medical charts and radio-
graphs were extensively reviewed by an independent
observer (J.R. Sangala). Patients included in this group had
multi-level symptomatic degenerative disc disease, disc
herniation, or stenosis of the cervical spine with or without
myelopathy and/or radiculopathy. All patients were care-
fully selected, having significant neural compression
extending beyond two or more disc spaces in the cervical
spine and were symptomatic even after appropriate con-
servative management. Patients who underwent surgery for
infection, neoplasm, trauma, or ossified posterior longitu-
dinal ligament (OPLL) were excluded from the study.
Patients were required to discontinue the use of all forms
of nicotine for at least 6 weeks prior to surgical interven-
tion. Serum nicotine levels were not measured
preoperatively. Mean age at the time of surgery was
47 years (range 30–75 years). A diagnosis of cervical
myelopathy was made in 8 out of 42 patients (19%) in the
ACDF group and 10 out of 38 patients (26%) in the ACCF
group. All patients underwent a preoperative clinical
evaluation along with magnetic resonance imaging and
radiographic studies of the cervical spine.
Operations were performed by the same neurosurgeon at
a single institution. Those patients with neural compression
from large end plate osteophytes at two adjacent levels
requiring resection of a larger portion of osseous material
were treated with a corpectomy. In those patients with
compression from disc material without significant end
plate osteophytes, the decision-making of the type of
decompression (either ACDF or ACCF) involved a detailed
discussion of potential benefits and risks of multi-level
ACDF and ACCF between the patient and surgeon. No
randomization was performed.
Based on the type of surgery the patients received, they
were divided into two groups: Group 1 consisted of
patients who received multi-level ACDF, and Group 2
consisted of patients who received single-/multi-level
ACCF. Titanium interbody cages and local autograft bone
were used in all procedures; specifically, Harms cages
(DePuy, Inc., Raynham, MA, USA) were used as structural
grafts during ACCF procedures and Rabea cages (Newport
Medical, Chanhassen, MN, USA) were used as interbody
grafts during ACDF. Cages were packed with autograft
from the resected vertebra in the corpectomy group and
from local cancellous bone harvested from the sternal
manubrium and/or local osteophytes in the discectomy
group. All cases were stabilized with a fixed ACP (DePuy,
Inc. or Synthes, Davos, Switzerland). The Rabea cage is a
‘‘box’’-type cage designed for implantation into the
cervical disc space after discectomy. It is designed to
facilitate fusion by use of forced friction to stabilize the
bones until osseous fusion both in and around the cage
occurred. Harms cages used in this study were single oval,
cylindrical meshed titanium cages. All the ACPs used in
this study were non-constrained.
Follow-up radiographs and clinical examinations were
obtained by the surgeon (F.L. Vale). Postoperative follow-
up visits were done regularly at 6 and 12 weeks, 6 months,
1 year, and 2 years (if necessary). All patients had ante-
rior–posterior and lateral cervical spine X-rays at each
visit; flexion/extension views were obtained after
12 weeks. Patients were followed for at least 12 months.
Fusion was judged by the absence of motion more than
2 mm between the spinous processes on flexion-extension
lateral radiographs, the absence of radiolucent gap between
the graft and end plate, and the presence of continuous
bridging trabeculae at the graft and end plate junction [51].
Radiographs were studied to look for hardware failure.
Subsidence and settling were recorded. Subsidence of more
than 3 mm was considered significant [18]. Cervical sag-
ittal balance (lordosis) was measured by Ishihara’s index.
A negative value of Ishihara’s index indicates kyphosis and
a positive value indicates lordosis [49]. Computerized
tomography and SPECT scans were not routinely obtained
to evaluate bone fusion. Postoperative clinical evaluation
of patients was based on Odom’s criteria (Table 2). All
patients had a minimum of 12 months of follow-up (range
12–46 months). None of the patients were lost to follow-
up.
Data were recorded and statistical analysis performed
using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA).
Data are presented as mean ± standard deviation. Mann–
Whitney rank sum and Chi-Square tests were used to
analyze differences in preoperative clinical and demo-
graphic characteristics (age, sex ratio) and in clinical
outcome variables between groups (Odom criteria, fusion
rates). Fisher exact test was used to analyze differences in
Table 2 Odom’s criteria
Rating Odom’s criteria
Excellent No complaints to cervical disc disease;
able to continue daily occupation
without impairment
Good Intermittent discomfort related to cervical
disease but not significantly interfering
with work
Satisfactory Subjective improvement but physical
activities limited
Poor No improvement or worse compared with
the condition before the operation
656 Eur Spine J (2009) 18:654–662
123

4. fusion and implant-related complications between groups.
Statistical significance was set at P 0.05.
Surgical techniques
Standard surgical techniques were used for both ACDF
and ACCF as previously described [29, 46]. In all
patients, ten pounds of traction was used to help stabilize
the cervical spine and provide intra-operative disc space
distraction. The Smith–Robinson technique was used for
Group 1 [46]. After confirmation and exposure of the
appropriate vertebral levels, a discectomy was performed
and a high-speed burr used to remove the cartilaginous
end plates from the adjoining vertebral bodies; excessive
removal of the subchondral bone was avoided. The
posterior longitudinal ligament was also completely
removed in all patients. Among 42 patients who under-
went ACDF, 38 patients underwent fusion at two levels,
while another 4 patients were fused at three levels
(Table 3). An ACP was used in all cases for internal
fixation.
For ACCF cases, the vertebral body was removed using
a channel technique [29]. In general, the anterior two-thirds
of the vertebral body was removed with rongeurs and the
posterior one-third was removed using a high-speed burr.
The posterior longitudinal ligament was completely
removed in all patients. An ACP was used routinely for
internal fixation. Among 38 patients who underwent
ACCF, corpectomy at one level was performed in 32
patients and two levels in 6 patients (Table 3).
The nuances mentioned by Perez-Cruet et al. [37] were
followed in both groups. All patients wore an Aspen cer-
vical collar for immobilization for at least 6 weeks
postoperatively.
Results
Group 1 was composed of 42 patients and Group 2 had 38
patients. The average age of patients in Group 1 was
46.2 years (range 35–60 years) and 50.1 years (range 35–
70 years) in Group 2 (Table 4). Outcome as measured by
Odom’s criteria was identical in both groups. Excellent
outcome was reported by 83.3% of the patients in Group 1
and 79% in Group 2 (Table 5). The average operative time
in Group 1 was 220 ± 30 min (range 160–280 min), and
was significantly higher (P 0.001) than Group 2, which
was 160 ± 20 min (range 110–230 min). The average
blood loss in Group 1 was 150 ± 23 ml (range 100–
350 ml), and was significantly lower (P 0.001) than
Group 2, which was 375 ± 30 ml (range 250–600 ml;
Fig. 1). None of the patients in either group developed any
new neurological deficits. No patient in either group
developed recurrent laryngeal nerve palsy, esophageal
injury, cerebrospinal fluid leak, or infections (Table 6).
Group 1 had one occurrence of pseudarthrosis and no
hardware failures. The patient with pseudarthrosis in Group
1 did not require re-operation and continued to be
asymptomatic at 5-year follow-up. In Group 2, two patients
developed pseudarthrosis and one patient had early hard-
ware failure. All three of these patients in Group 2 required
re-operation. The fusion rates in Group 1 were 97.6 and
92.1% for Group 2. Sagittal balance was well-maintained
(80%) in both groups.
The case of early hardware failure in Group 2 was
associated with screw pull-out and cage dislodgement, and
the patient presented with dysphagia. This patient was
managed by removal of the ACP, placing a buttress ante-
rior plate, repositioning the cage, and posterior fusion. The
two patients with pseudarthrosis in Group 2 were managed
by posterior fusion.
Table 3 Showing number of disc levels operated in both the groups
Disc levels Group1 (ACDF)
n = 42
Group 2 (ACCF)
n = 38
C3–4, C4–5 0 1
C3–4, C4–5, C5–6 0 1
C4–5, C5–6 6 9
C4–5, C5–6, C6–7 4 5
C5–6, C6–7 32 22
Table 4 Population demographics
Group 1
(ACDF)
Group 2
(ACCF)
Number 42 38
No of females/no of patients 21 17
Average age (years) 46.2 50
Myelopathy as pre-dominant presentation 7 11
Average follow up (years) 2.3 2.2
Average operation time (min) 220 160
Average blood loss (ml) 150 375
Table 5 Results of surgery according to Odom’s criteria sorted by
procedure
Excellent Good Fair Poor
Group1 (ACDF) n = 42 35 (83.3%) 6 (14.3%) 1 (2.34%) 0 (0%)
Group2 (ACCF) n = 38 30 (78.9%) 6 (16.8%) 2 (5.3%) 0 (0%)
There is no statistically significant difference between the two groups
Eur Spine J (2009) 18:654–662 657
123

5. Discussion
Surgical treatment of multi-level cervical spondylosis is
challenging. Type of decompression and the technique of
reconstruction are the two important decisions to be made.
The types of decompression usually used are multi-level
ACDF or single-/multi-level ACCF. Since its initial
description in the mid–1950 s by Smith and Robinson [46]
and Cloward [11], anterior cervical discectomy for cervical
spondylosis has been reported with good clinical outcome
and fusion rates. In the decades that followed, various
modifications of these landmark techniques have been
described in the literature, including the addition of ACP
by Orozco in 1970 [34].
Fusion rates are high (up to 96%) using either an auto-
graft or allograft with ACP for single-level ACDF [39].
However, fusion rates were much lower for multi-level
ACDF using either autograft or an allograft. Non-union
rates of 26% for two-level ACDF and 48% for three-level
ACDF have been reported in the literature [48]. One of the
reasons postulated for the high non-union rates after multi-
level ACDF is the higher number of fusion surfaces
involved. Some surgeons have advocated the use of cer-
vical corpectomies as an alternative to multiple interbody
grafts, citing a decreased rate of pseudarthrosis secondary
to fewer graft–host interfaces where fusion needs to occur
[33, 48, 50]. In anterior reconstruction after corpectomy,
fibular strut grafts were associated with a non-union rate as
high as 41% for allograft and 27% for autograft [16]. Some
studies have reported high fusion rates using an autograft
for reconstruction after corpectomy [20]; however, the use
of autograft has significant donor site morbidity [43]. Also,
results of using autograft for vertebral defects greater than
6 cm have been disappointing [48]. Hence, the use of
autograft for reconstruction is decreasing. Allograft has the
advantage of providing immediate mechanical strength,
availability in the desired shape, size, and quantity. But the
important disadvantage of allograft is the high non-union
rate when used for multi-level ACDF and also corpectomy
[46]. In view of the disadvantages of both allograft and
autograft, the TMC designed by Harms in 1986 became an
alternative method of reconstruction in these procedures.
Titanium mesh cages have been used in anterior
reconstruction of the spine for trauma and tumors with
adequate results [29, 50]. High fusion rates have been
reported after cervical corpectomies using TMC (up to
100%) [50]. All of the studies we reviewed have reported
higher fusion rates after corpectomy than multi-level
ACDF [33, 48, 61] when using similar reconstruction
techniques (Table 1). Contrary to these results, our study
found slightly higher fusion rates after multi-level ACDF
than ACCF. We believe these findings are due to the higher
fusion rates in the ACDF group, since the fusion rates for
the ACCF group (92.1%) were comparable to other series
Fig. 1 a Graph depicting the lower average blood loss in the ACDF
group. b Graph depicting the lower operative time in the ACCF group
Table 6 Complications
Complications Group1 Group2
Early graft displacement 0 1
Pseudarthrosis 1 2
Infections 0 0
Recurrent laryngeal nerve injury 0 0
Esophageal injuries 0 0
CSF leaks 0 0
658 Eur Spine J (2009) 18:654–662
123

6. [29, 50]. The higher fusion rates in the ACDF group in our
study could be due to strict adherence to surgical principles
of anterior cervical fusion, such as preserving the end
plates, avoiding over-distraction, the use of local autograft
and supplemental autograft from sternal manubrium in
selected cases, and the use of cages instead of allograft.
Our results are similar to those of Wang et al. [59], who
reported similar fusion rates in both multi-level ACDF and
ACCF groups using allograft.
Bone graft is usually packed inside the cage to serve as a
bone graft substrate. In our study, the cages were filled with
local bone saved from corpectomy in Group 2 and from
osteophytes and sternal manubrium in Group 1. Sternal
manubrium has shown to be a safe and effective source of
autograft without donor site morbidity [36].
We believe that our study is unique in that it presents for
the first time the outcome of multi-level ACDF and ACCF
using TMCs for reconstruction. Similar studies have been
performed using either an allograft or autograft (Table 1).
Other contemporary methods used for reconstruction
include artificial disc, PEEK cages, carbon fiber cages, and
titanium rods with buttress prosthesis [5]. The cervical
artificial disc has been recently approved by the United
States Food and Drug Administration (FDA) and is still not
widely used [27]. PEEK cages have certain advantages
over metal cages as they are radiolucent and have the same
modulus of elasticity as bone [26] and appear to be an
alternative to metal cages. Numerous modifications have
been used for corpectomy, reporting greater safety and
equivalent outcome [9, 21, 60]. An important advance in
spinal fusion has been the use of BMP-2. However, the
FDA has not approved the use of BMP-2 for cervical
fusion, and complications have been reported with the use
of BMP-2 in cervical fusion [54]. Polymethylmethacrylate
(PMMA) has been reported to produce satisfactory results
in ACDF [10]. However, PMMA fails to meet the
requirements for an interbody fusion device and causes
necrosis of adjacent vertebrae [56].
The strengths of our study include the involvement of
two groups with very similar demographics, a relatively
large number of patients, use of a contemporary technique
for spinal reconstruction, all surgeries performed by a
single surgeon at a single institution, and good follow-up
data for at least 1 year post-surgery (no patients lost to
follow-up). The limitations of our study include the
potential for surgeon bias in patient selection in patients
with no compelling reason to perform a corpectomy, lack
of randomization of the treatment group, and importantly,
the disadvantages of a retrospective study.
In our series, no patient experienced infection, recurrent
laryngeal nerve palsy, or esophageal perforations. Infection
rates in most series are low (0.5%) [31, 52, 63]. Although
no patient in our series developed cerebrospinal fluid leak,
several published reports describe this problem in corpec-
tomies and were associated with severe spondylosis,
multiple re-operations, or OPLL [14, 15, 28].
Clinical outcome was comparable in both groups and
was similar to earlier reports [59]. However, the rate of
postoperative hardware failure and fusion failure was
higher in those undergoing corpectomy. Our study under-
lines both the advantages and the pitfalls with the use of
TMCs and ACPs in reconstruction of the anterior cervical
spine. The advantages are the high fusion rates as shown in
the present study. These high fusion rates could be due to
the decreased incidence of significant subsidence. Although
Gercek et al. [18] reported a high incidence of subsidence,
other authors with larger numbers of patients have reported
lower rates [4]. We believe that careful preservation of the
end plates and avoiding of over-distraction are very
important to avoid significant subsidence. Non-significant
subsidence is not associated with any complications and
may, on the other hand, help fusion [42]. TMCs are suffi-
ciently rigid and are designed to resist compression and
torque [35]. The addition of an ACP resists bending forces
and theoretically protects against dislodgement of the cage,
helps in maintaining cervical lordosis [24, 53], and is
reported to have higher fusion rates [58, 62].
Though it has become a routine practice to use ACP after
multi-level ACDF and also single-/multi-level corpectomy,
there are both clinical and biomechanical disadvantages of
using plates. One disadvantage of using a fixed ACP is that
it prevents subsidence and may prevent the proper contact
of the graft with vertebral end plates [47]. The use of ACP
in spine reconstruction with multi-level corpectomies is
also technically demanding and fraught with complications
(e.g., plate migration or dislodgment) [12, 41]. DiAngelo
et al. [13] demonstrated that, following multiple level
corpectomies and strut grafting, though ACP effectively
increased stiffness and decreased local cervical motion
after corpectomy, it was also shown to reverse graft loads
and excessively load the graft in extension, which may
promote pistoning and failure of a multi-level construct.
Similarly, Foley et al. [17] in a biomechanical cadaveric
study concluded that, although multi-level cervical instru-
mentation effectively increased stiffness after corpectomy,
anterior or posterior plating alone excessively loads the
graft with a small degree of motion and may promote
pistoning and failure of multi-level constructs.
As described by Vaccaro et al. [55] and Sasso et al. [41],
the incidence of early plate/construct failure increases
significantly with a three-level corpectomy compared with
one or two levels. In their practice, the authors limit
corpectomy procedures to two levels. In our series, in
Group 2, we used an end-construct plate fixation that
spanned the graft. Our study also highlights the pitfalls of a
long fusion ending at C7, which has also been reported
Eur Spine J (2009) 18:654–662 659
123

7. earlier [57]. In addition to graft length, the actual level
fused appear to have an effect on graft stability. All 3
complications in the corpectomy group involved C6 as the
caudal corpectomy level, with the fusion extending to the
C7 vertebral body. A large moment arm can be generated at
the ends of the construct, potentially leading to plate dis-
lodgment. The tendency for bone fusion failure for cervical
procedures extending to the level of C7 might be related to
transition from cervical lordosis to thoracic kyphosis. This
sharp angular change could lead to increased stress at the
graft end plate interface, resulting in higher probability of
construct failure, as described by others [19]. We are also
aware that this inherent morbidity of multi-level fusions
can simply reflect the fact that long constructs usually do
extend to C7. Consideration of supplemental dorsal fixation
and/or external immobilization (halo vest) should be given
for a long construct that spans the cervical–thoracic junc-
tion (C7–T1). We believe that the increased rigidity
afforded by posterior segmental fixation may decrease the
likelihood of cage dislodgment observed in the setting of
long fusion with anterior instrumentation alone. The
increased fixation provided by dorsal fusion allows the
large-cantilever forces generated across the cervical spine
to be distributed across multiple screw–host interfaces.
The results of our study indicate there is less morbidity
with multi-level ACDF than ACCF. Some biomechanical
studies have also addressed the stability of multi-level
ACDF and ACCF [25, 44, 45]. In a biomechanical
cadaveric study, Singh et al. [44] evaluated the biome-
chanical stability of ACP fixation following three strategies
for decompression of spondylosis (three levels) of the
cervical spine: three-level discectomy, single-level corp-
ectomy, and two-level corpectomy. They concluded that a
large moment arm is generated at the ends of the construct
after multi-level corpectomy, potentially leading to plate
migration or dislodgement. They also reported that seg-
mental plate fixation affords a more biomechanically rigid
reconstruction. In another biomechanical cadaveric study,
Kirkpatrick et al. [25] determined the biomechanical
behavior of the cervical spine after multi-level corpectomy
and reconstruction with a strut graft and supplementation
with anterior and posterior plates. They concluded that
plating increases stability of the spine after multi-level
corpectomy. Interestingly, they found that posterior plating
provided more rigidity than ACP. These two cadaveric
studies appear to be the biomechanical basis of clinical
findings reported in our study.
Conclusions
On the basis of our findings, we believe that both multi-
level ACDF and ACCF with anterior cervical
reconstruction with a TMC filled with autograft and ACP
for treating multi-level cervical spondylosis have high
fusion rates and good clinical outcome. However, there is a
higher rate of early hardware failure and pseudarthrosis
after ACCF than ACDF. In view of the equivalent fusion
rates using TMC and autograft, in the absence of specific
pathology that would require removal of vertebral body,
multi-level ACDF using interbody cages and autologous
bone graft could result in lower morbidity. We also propose
that in patients who undergo a multi-level corpectomy with
a construct that involves C7, additional posterior stabi-
lization or postoperative immobilization should be
considered to protect against construct failure.
Acknowledgments The authors acknowledge Katheryne Downes,
M.PÁH., for her assistance in statistical data analysis. No external
sources of funding. No commercial interest in any of the devices
mentioned in the article by any of the authors.
References
1. An HS, Simpson JM, Glover JM et al (1995) Comparison
between allograft plus demineralized bone matrix versus auto-
graft in anterior cervical fusion. A prospective multicenter study.
Spine 20:2211–2216
2. Bishop RC, Moore KA, Hadley MN (1996) Anterior cervical
interbody fusion using autoeneic and allogeneic bone graft sub-
strate: a prospective comparative analysis. J Neurosurg 85:206–210
3. Bolesta MJ, Rechtine GRII, Chrin AM (2000) Three- and four-
level anterior cervical discectomy and fusion with plate fixation:
a prospective study. Spine 25:2040–2046. doi:10.1097/0000
7632-200008150-00007
4. Borm W, Seitz K (2004) Use of cervical stand-alone cages. Eur
Spine J 13:474–477
5. Brazenor GA (2007) Comparison of multisegment anterior cer-
vical fixation using bone strut graft versus a titanium rod and
buttress prosthesis: analysis of outcome with long-term follow-up
and interview by independent physician. Spine 32:63–71. doi:
10.1097/01.brs.0000250304.24001.24
6. Brown MD, Malinin TI, Davis PB (1976) A roentgenographic
evaluation of frozen allografts versus autografts in anterior cer-
vical spine fusions. Clin Orthop Relat Res 119:231–236
7. Cauthen JC, Kinard RE, Vogler JB et al (1998) Outcome analysis
of noninstrumented anterior cervical discectomy and interbody
fusion in 348 patients. Spine 23:188–192. doi:10.1097/00007632-
199801150-00008
8. Cauthen JC, Theis RP, Allen AT (2003) Anterior cervical fusion:
a comparison of cage, dowel and dowel-plate constructs. Spine J
3:106–117. doi:10.1016/S1529-9430(02)00533-8
9. Chacko AG, Daniel RT (2007) Multilevel cervical oblique
corpectomy in the treatment of ossified posterior longitudinal
ligament in the presence of ossified anterior longitudinal liga-
ment. Spine 32:E575–E580
10. Chen JF, Wu CT, Lee SC et al (2005) Use of a polymethyl-
methacrylate cervical cage in the treatment of single-level
cervical disc disease. J Neurosurg Spine 3:24–28
11. Cloward RB (1958) The anterior approach for removal of rup-
tured cervical disks. J Neurosurg 15:602–617
12. Daubs MD (2005) Early failures following cervical corpectomy
reconstruction with titanium mesh cages and anterior plating.
Spine 30:1402–1406. doi:10.1097/01.brs.0000166526.78058.3c
660 Eur Spine J (2009) 18:654–662
123

8. 13. DiAngelo DJ, Foley KT, Vossel KA et al (2000) Anterior cervical
plating reverses load transfer through multilevel strut-grafts.
Spine 25:783–795. doi:10.1097/00007632-200004010-00005
14. Emery SE, Bohlman HH, Bolesta MJ et al (1998) Anterior cer-
vical decompression and arthrodesis for the treatment of cervical
spondylotic myelopathy. Two to seventeen-year follow-up. J
Bone Joint Surg Am 80:941–951. doi:10.1302/0301-620X.
80B6.9517
15. Epstein NE (2000) Anterior cervical diskectomy and fusion
without plate instrumentation in 178 patients. J Spinal Disord
13:1–8. doi:10.1097/00002517-200002000-00001
16. Fernyhough JC, White JI, LaRocca H (1991) Fusion rates in
multilevel cervical spondylosis comparing allograft fibula with
autograft fibula in 126 patients. Spine 16:S561–S564
17. Foley KT, DiAngelo DJ, Rampersaud YR et al (1999) The in
vitro effects of instrumentation on multilevel cervical strut-graft
mechanics. Spine 24:2366–2376. doi:10.1097/00007632-1999
11150-00014
18. Gercek E, Arlet V, Delisle J et al (2003) Subsidence of stand-
alone cervical cages in anterior interbody fusion: warning. Eur
Spine J 12:513–516. doi:10.1007/s00586-003-0539-6
19. Gonugunta V, Krishnaney AA, Benzel EC (2005) Anterior cer-
vical plating. Neurol India 53:424–432
20. Gore DR, Sepic SB (1984) Anterior cervical fusion for degen-
erated or protruded discs. A review of one hundred forty-six
patients. Spine 9:667–671. doi:10.1097/00007632-198410000-
00002
21. Groff MW, Sriharan S, Lee SM et al (2003) Partial corpectomy
for cervical spondylosis. Spine 28:14–20. doi:10.1097/00007632-
200301010-00005
22. Hacker RJ, Cauthen JC, Gilbert TJ et al (2000) A prospective
randomized multicenter clinical evaluation of an anterior cervical
fusion cage. Spine 25:2646–2655. doi:10.1097/00007632-2000
10150-00017
23. Hwang SL, Lin CL, Lieu AS et al (2004) Three-level and four-
level anterior cervical discectomies and titanium cage-augmented
fusion with and without plate fixation. J Neurosurg Spine 1:160–
167
24. Katsuura A, Hukuda S, Imanaka T et al (1996) Anterior cervical
plate used in degenerative disease can maintain cervical lordosis.
J Spinal Disord 9:470–476. doi:10.1097/00002517-199612000-
00003
25. Kirkpatrick JS, Levy JA, Carillo J et al (1999) Reconstruction
after multilevel corpectomy in the cervical spine. A sagittal plane
biomechanical study. Spine 24:1186–1191. doi:10.1097/0000
7632-199906150-00003
26. Liao JC, Niu CC, Chen WJ et al. (2007) Polyetheretherketone
(PEEK) cage filled with cancellous allograft in anterior cervical
discectomy and fusion. Int Orthop Epub ahead of print
27. Lind BI, Zoega B, Rosen H (2007) Autograft versus interbody
fusion cage without plate fixation in the cervical spine: a ran-
domized clinical study using radiostereometry. Eur Spine J
16:1251–1256. doi:10.1007/s00586-007-0337-7
28. Macdonald RL, Fehlings MG, Tator CH et al (1997) Multilevel
anterior cervical corpectomy and fibular allograft fusion for
cervical myelopathy. J Neurosurg 86:990–997
29. Majd ME, Vadhva M, Holt RT (1999) Anterior cervical recon-
struction using titanium cages with anterior plating. Spine
24:1604–1610. doi:10.1097/00007632-199908010-00016
30. McAfee PC, Bohlman HH, Ducker TB et al (1995) One-stage
anterior cervical decompression and posterior stabilization. A
study of one hundred patients with a minimum of two years of
follow-up. J Bone Joint Surg Am 77:1791–1800
31. Metzger CS, Schlitt M, Quindlen EA et al (1989) Small central
cervical disc syndrome: evaluation and treatment of chronic
disabling neck pain. J Spinal Disord 2:234–237. doi:
10.1097/00002517-198912000-00004
32. Moreland DB, Asch HL, Clabeaux DE et al (2004) Anterior
cervical discectomy and fusion with implantable titanium cage:
initial impressions, patient outcomes and comparison to fusion
with allograft. Spine J 4:184–191. doi:10.1016/j.spinee.2003.
05.001
33. Nirala AP, Husain M, Vatsal DK (2004) A retrospective study of
multiple interbody grafting and long segment strut grafting fol-
lowing multilevel anterior cervical decompression. Br J
Neurosurg 18:227–232. doi:10.1080/02688690410001732643
34. Omeis I, DeMattia JA, Hillard VH et al (2004) History of
instrumentation for stabilization of the subaxial cervical spine.
Neurosurg Focus 16:E10. doi:10.3171/foc.2004.16.1.11
35. Orr RD, Zdeblick TA (1999) Cervical spondylotic myelopathy.
Approaches to surgical treatment. Clin Orthop Relat Res 359:58–
66. doi:10.1097/00003086-199902000-00007
36. Peelle MW, Rawlins BA, Frelinghuysen P (2007) A novel source
of cancellous autograft for ACDF surgery: the manubrium. J
Spinal Disord Tech 20:36–41. doi:10.1097/01.bsd.0000211230.
89554.5c
37. Perez-Cruet MJ, Samartzis D, Fessler RG (2006) Anterior cer-
vical discectomy and corpectomy. Neurosurgery 58:355–359.
doi:10.1227/01.NEU.0000205285.20336.C2
38. Rish BL, McFadden JT, Penix JO (1976) Anterior cervical fusion
using homologous bone grafts: a comparative study. Surg Neurol
5:119–121
39. Samartzis D, Shen FH, Goldberg EJ et al (2005) Is autograft the
gold standard in achieving radiographic fusion in one-level
anterior cervical discectomy and fusion with rigid anterior plate
fixation? Spine 30:1756–1761. doi:10.1097/01.brs.0000172148.
86756.ce
40. Samartzis D, Shen FH, Matthews DK et al (2003) Comparison of
allograft to autograft in multilevel anterior cervical discectomy
and fusion with rigid plate fixation. Spine J 3:451–459. doi:
10.1016/S1529-9430(03)00173-6
41. Sasso RC, Ruggiero RA Jr, Reilly TM et al (2003) Early
reconstruction failures after multilevel cervical corpectomy.
Spine 28:140–142. doi:10.1097/00007632-200301150-00009
42. Schmieder K, Wolzik-Grossmann M, Pechlivanis I et al (2006)
Subsidence of the wing titanium cage after anterior cervical
interbody fusion: 2-year follow-up study. J Neurosurg Spine
4:447–453. doi:10.3171/spi.2006.4.6.447
43. Silber JS, Anderson DG, Daffner SD et al (2003) Donor site
morbidity after anterior iliac crest bone harvest for single-level
anterior cervical discectomy and fusion. Spine 28:134–139. doi:
10.1097/00007632-200301150-00008
44. Singh K, Vaccaro AR, Kim J et al (2003) Biomechanical com-
parison of cervical spine reconstructive techniques after a
multilevel corpectomy of the cervical spine. Spine 28:2352–2358.
doi:10.1097/01.BRS.0000085344.22471.23
45. Singh K, Vaccaro AR, Kim J et al (2004) Enhancement of sta-
bility following anterior cervical corpectomy: a biomechanical
study. Spine 29:845–849. doi:10.1097/00007632-200404150-
00005
46. Smith GW, Robinson RA (1958) The treatment of certain cer-
vical-spine disorders by anterior removal of the intervertebral
disc and interbody fusion. J Bone Joint Surg Am 40-A:607–624
47. Stewart TJ, Schlenk RP, Benzel EC (2007) Multiple level disc-
ectomy and fusion. Neurosurgery 60:S143–S148. doi:10.1227/01.
NEU.0000217015.96212.1B
48. Swank ML, Lowery GL, Bhat AL et al (1997) Anterior cervical
allograft arthrodesis and instrumentation: multilevel interbody
grafting or strut graft reconstruction. Eur Spine J 6:138–143. doi:
10.1007/BF01358747
Eur Spine J (2009) 18:654–662 661
123

9. 49. Takeshita K, Murakami M, Kobayashi A et al (2001) Relation-
ship between cervical curvature index (Ishihara) and cervical
spine angle (C2–7). J Orthop Sci 6:223–226. doi:10.1007/
s007760100038
50. Thalgott JS, Xiongsheng C, Giuffre JM (2003) Single stage
anterior cervical reconstruction with titanium mesh cages, local
bone graft, and anterior plating. Spine J 3:294–300. doi:10.1016/
S1529-9430(02)00588-0
51. Thome C, Krauss JK, Zevgaridis D (2004) A prospective clinical
comparison of rectangular titanium cages and iliac crest auto-
grafts in anterior cervical discectomy and fusion. Neurosurg Rev
27:34–41. doi:10.1007/s10143-003-0297-2
52. Thorell W, Cooper J, Hellbusch L et al (1998) The long-term
clinical outcome of patients undergoing anterior cervical disc-
ectomy with and without intervertebral bone graft placement.
Neurosurgery 43:268–274. doi:10.1097/00006123-199808000-
00050
53. Troyanovich SJ, Stroink AR, Kattner KA et al (2002) Does
anterior plating maintain cervical lordosis versus conventional
fusion techniques? A retrospective analysis of patients receiving
single-level fusions. J Spinal Disord Tech 15:69–74
54. Tumialan LM, Rodts GE (2007) Adverse swelling associated
with use of rh-BMP-2 in anterior cervical discectomy and fusion.
Spine J 7:509–510. doi:10.1016/j.spinee.2007.03.005
55. Vaccaro AR, Falatyn SP, Scuderi GJ et al (1998) Early failure of
long segment anterior cervical plate fixation. J Spinal Disord
11:410–415
56. van den Bent MJ, Oosting J, Wouda EJ et al (1996) Anterior
cervical discectomy with or without fusion with acrylate. A
randomized trial. Spine 21:834–840. doi:10.1097/00007632-
199604010-00011
57. Wang JC, Hart RA, Emery SE et al (2003) Graft migration or
displacement after multilevel cervical corpectomy and strut
grafting. Spine 28:1016–1022. doi:10.1097/00007632-20030
5150-00011
58. Wang JC, McDonough PW, Endow KK et al (2000) Increased
fusion rates with cervical plating for two-level anterior cervical
discectomy and fusion. Spine 25:41–45. doi:10.1097/00007632-
200001010-00009
59. Wang JC, McDonough PW, Endow KK et al (2001) A compar-
ison of fusion rates between single-level cervical corpectomy and
two-level discectomy and fusion. J Spinal Disord 14:222–225.
doi:10.1097/00002517-200106000-00006
60. Ying Z, Xinwei W, Jing Z et al (2007) Cervical corpectomy with
preserved posterior vertebral wall for cervical spondylotic mye-
lopathy: a randomized control clinical study. Spine 32:1482–
1487. doi:10.1097/BRS.0b013e318068b30a
61. Yonenobu K, Fuji T, Ono K et al (1985) Choice of surgical
treatment for multisegmental cervical spondylotic myelopathy.
Spine 10:710–716. doi:10.1097/00007632-198510000-00004
62. Zaveri GR, Ford M (2001) Cervical spondylosis: the role of
anterior instrumentation after decompression and fusion. J Spinal
Disord 14:10–16. doi:10.1097/00002517-200102000-00003
63. Zdeblick TA, Bohlman HH (1989) Cervical kyphosis and mye-
lopathy. Treatment by anterior corpectomy and strut-grafting. J
Bone Joint Surg Am 71:170–182
662 Eur Spine J (2009) 18:654–662
123